How does a credible voice sound?

  • Credibility is a crucial social factor that influences people's perception and decision-making. This study explores the acoustic attributes that define credible speech compared to neutral and ironic speech. A custom-built German corpus was developed, containing speech samples recorded from amateurs to enhance ecological validity. The study extracted a broad set of audio features from these samples, employing recursive feature elimination to identify the most influential attributes. These were then analyzed using a machine-learning-supported multinomial logistic regression model. The results indicated significant differences in the acoustic features associated with credible speech compared to neutral and ironic speech. Key findings include the role of a higher energy level (1st mel-frequency cepstral coefficient) in credible compared to neutral and ironic speech and a higher speaking rate in both credible and ironic compared to neutral speech. Also, irony is characterized by moreCredibility is a crucial social factor that influences people's perception and decision-making. This study explores the acoustic attributes that define credible speech compared to neutral and ironic speech. A custom-built German corpus was developed, containing speech samples recorded from amateurs to enhance ecological validity. The study extracted a broad set of audio features from these samples, employing recursive feature elimination to identify the most influential attributes. These were then analyzed using a machine-learning-supported multinomial logistic regression model. The results indicated significant differences in the acoustic features associated with credible speech compared to neutral and ironic speech. Key findings include the role of a higher energy level (1st mel-frequency cepstral coefficient) in credible compared to neutral and ironic speech and a higher speaking rate in both credible and ironic compared to neutral speech. Also, irony is characterized by more high-frequency content (mean spectral centroid) compared to credible or neutral speech. Gender differences in spoken irony involve a greater influence of speaking rate in women's speech, while high pitch plays a more significant role in men's speech. This research thus contributes to the understanding of how credibility is conveyed through speech and offers insights for applications in communication, media, and artificial intelligence. The study also highlights the methodological advancements made by incorporating a diverse range of acoustic features and employing a robust machine-learning framework.show moreshow less

Download full text files

Export metadata

Metadaten
Author:Jochen SteffensORCiD, Patrick Blättermann, Maximilian Sattler, Kevin TangORCiD
Qualitätssicherung:peer reviewed
open access:Hybrid - Subskriptionszeitschrift
Institutes:Institute of Sound and Vibration Engineering (ISAVE)
Research fields:Kognition und Wissen / Sprache und Spracherlernen
Fachbereich/Einrichtung:Hochschule Düsseldorf / Fachbereich - Medien
Document Type:Article
Year of Completion:2025
Language of Publication:English
Publisher:Acoustical Society of America
Parent Title (English):The Journal of the Acoustical Society of America
Volume:157
Issue:5
Page Number:14
First Page:3780
Last Page:3792
URN:urn:nbn:de:hbz:due62-opus-53133
DOI:https://doi.org/10.1121/10.0036646
ISSN:1520-8524
GND Keyword:Sprechen; Glaubwürdigkeit; Akustik
Corresponding Author:Jochen Steffens, Kevin Tang
Information on the Research Data:The data that support the findings of this study are available within the article and its supplementary material and are openly available in the Open Science Framework (OSF) repository at https://doi.org/10.17605/OSF.IO/MTCUH
Licence (German):Creative Commons - CC BY - Namensnennung 4.0 International
Release Date:2025/05/21
Diese Webseite verwendet technisch erforderliche Session-Cookies. Durch die weitere Nutzung der Webseite stimmen Sie diesem zu. Unsere Datenschutzerklärung finden Sie hier.